The art of surveying: what works (and what doesn’t)

For communication professionals, the survey is an extremely valuable – indeed, often indispensable – tool, offering a unique window of insight onto a particular demographic, audience or customer base.

In order to fully-harness the power of the survey, however, it is important to select your questions carefully.  Get it wrong and what could have been a potentially invaluable piece of market research to add to the PR pro’s portfolio becomes nothing short of useless.

With this in mind, let’s begin by looking at the types of questions:

1.       Factual

Questions where people know the answer without thinking. For example: age, gender, family composition, main supermarket, which high profile TV programmes they watch etc.

2.       Simple

For example, providing people with two, three or even more options and ask which they prefer (asking why and/or rating them is harder than just picking a winner).

3.       Time Sensitive

Questions about things that are easy to remember within set parameters. For example, “Did you buy a car this year?” or “Did you go to the movies last week?”

4.       Rough/Average

Questions that can be made easy to answer by grouping the answers. For example, “How many times do you eat in a restaurant in a typical month?” With answers perhaps grouped as:  “Never,” “One or two times a month,” “Three to five times a month,” “More than five times a month.”

It seems straightforward, but it can be all too easy to word a question in a way that confuses the respondent and causes incomplete responses. Some examples of these sorts of questions are:

1.       Unspecific

Not putting a date range or asking the respondent to think too far back – “How many times do you eat ice cream? Per day, per week, per month?”

2.       The double-barrelled question

“Was the train clean and on time?” This invalidates a question; if the train was clean but late, the respondent will be stuck.

3.       Socially loaded questions

“How often do you feed junk food to your children?” These questions can be made less judgemental by not using loaded terms such as “junk food” and instead showing parents a list of foods (including, say, crisps and burgers, but also apples and carrots) and ask which their child eats weekly.

In addition, respondents may not know the answer to all questions. So, for questions like “What is the speed of the internet connection to your home?” or “Will you vote in the next local election?, it is important to include an “I don’t know” option. Sometimes the number of those not knowing the response can be an interesting finding in itself.

Asking survey questions is a science and an art; it is not just an extension of how we ask questions in everyday life. It is worth thinking about what sorts of questions work, what sorts of questions don’t, and piloting your surveys to make sure they achieve exactly what you want them to.